Enhancing pronunciation training with artificial intelligence
Journal of Second Language Pronunciation · 02.11.2026 · 12 (3) · 432-437
DOI: 10.1075/jslp.26016.hay ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Second Language Pronunciation, online erschienen am 09.02.2026. Verfasst von Emma Hayter. Die Arbeit nennt 14 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
In many second language (L2) classrooms, English pronunciation receives limited instructional attention (Farhat & Dzakiria, 2017;McCrocklin, 2016).For this reason, autonomous learning may serve as an effective supplement to classroombased pronunciation training (McCrocklin, 2016).Rapid advancements in artificial intelligence (AI), including tools such as text-to-speech (TTS), chatbot-based interaction, and automatic speech recognition (ASR), have expanded the potential for self-directed language learning through mobile applications.One such app that has attracted recent interest is BoldVoice (Wellocution Inc., 2026), which is designed to help learners improve their American English pronunciation.As of January 15, 2026, BoldVoice is available to download on both iOS and Android devices.During the account setup process, the learner must answer a few questions to personalize their learning experience, such as indicating their pronunciation goals, English proficiency level, and first language.The answers provided in the preliminary setup stage help to customize the pronunciation lessons for the learner.BoldVoice offers a seven-day free trial upon completing the initial assessment; as the app is subscription-based, monthly Standard subscriptions start at $25 USD per month or $150 USD per year.Before starting the personalized lessons, an AI tool called SpeechScan assesses the learner's English pronunciation and identifies which phonemes they have difficulty pronouncing.Then, SpeechScan provides the learner with an evaluation of their proficiency level and identifies their top three pronunciation strengths (e.g., /p/, /d/, /k/), and top three areas needing improvement (e.g., /θ/, /ʃ/, /ɚ/).ASR grades the learner's pronunciation accuracy by assigning a Goodness of Pronunciation (GOP) score (Witt & Young, 2000).Correctly pronounced phonemes are highlighted in green, whereas incorrectly pronounced sounds appear in red (Figure 1).
Zitieren
APA
Hayter, E. (2026). Enhancing pronunciation training with artificial intelligence. Journal of Second Language Pronunciation, 12(3), 432-437. https://doi.org/10.1075/jslp.26016.hay
Vancouver
Hayter E. Enhancing pronunciation training with artificial intelligence. Journal of Second Language Pronunciation. 2026;12(3):432-437. doi:10.1075/jslp.26016.hay
BibTeX
@article{Hayter2026,
author = {Hayter, Emma},
title = {Enhancing pronunciation training with artificial intelligence},
journal = {Journal of Second Language Pronunciation},
year = {2026},
volume = {12},
number = {3},
pages = {432--437},
doi = {10.1075/jslp.26016.hay}
}
EUVIMED